MODELING AND CONTROL STRATEGIES FOR AUTONOMOUS VEHICLES AND TRUCK PLATOONS IN MIXED TRAFFIC SCENARIOS

Since the dawn of civilization, people started to move around and transport goods. The more in time we move forward, the more we find bigger cities and therefore bigger economies, and that is thanks to human creativity and inventions, all the way from the wheel to the transportation industry nowadays. Technology advanced and a person's well-being and comfort became more and more important. Unfortunately, the infrastructure and the environment could not keep up with all these advancements; problems could not be foreseen, so that is what led us to the challenges that we face today: pollution and environmental disasters, traffic congestion and accidents, energy resources scarcity and high price, etc. Every day, the challenges posed by economic globalization and climate change become more pressing for institutions and governments to tackle. In this scenario, the role of the transport sector, particularly the road network, is undeniably significant. The increasing demand for both passenger travel and goods transportation has contributed to the ongoing deterioration of quality of life. In addition, both routine and unexpected traffic jams continue to frustrate citizens, eroding public trust in the reliability of transportation systems. Researchers and institutions tried and succeeded in solving, even if partially, some of these problems, at least in some parts of the world: emission reduction in more effective fuel engines, traffic control methods, upgrade of infrastructure, campaigns of information and teaching to the public, etc. Like mentioned before, technology keeps advancing and the world keeps witnessing new inventions, like electric vehicles, autonomous driving and platooning. These advancements not only promise individual comfort but also solutions to the problems. For example, electric cars powered by renewable resources can reduce significantly green house emissions and pollution in cities. Autonomous electric vehicles present a great potential to optimize energy consumption; that's why their trajectories need to be carefully planned taking advantage of their regenerative brakes to get optimized speed profiles. Autonomous vehicles (AVs) in general can improve traffic flow and prevent congestion from forming, of course apart from the other benefits of more comfort, less travel time, energy efficiency and so on. They have as well the potential to significantly improve traffic flow around signalized intersections by leveraging real-time communication and advanced control systems. Equipped with Vehicle-to-Infrastructure (V2I) technology, AVs can receive traffic light timing information and adjust their speed accordingly to minimize stops and optimize fuel efficiency. This coordination reduces the formation of queues and prevents sudden acceleration or deceleration, leading to smoother traffic flow. This concept along with the vehicles being electric, fall in the domain of eco-driving, a sustainable concept in the automotive sector. The problem is that not all traditional cars can be immediately, or at least fast enough, be replaced by AVs; this creates mixed traffic scenarios that are not studied enough and do not completely provide the promised solution to congestion. Moreover, some challenges arise from the unpredictable behavior of human-driven vehicles (HVs) and the lack of vehicle-to-vehicle (V2V) and V2I communication. Modelling and controlling the interaction between the two types of vehicles is crucial, especially in the absence of real life experiments. Another subject related to mixed traffic is the travel of autonomous truck platoons on freeways. Platooning, especially for trucks, can reduce significantly the air drag for the vehicles traveling within the platoon. Other than fuel efficiency, autonomous truck platooning can improve traffic flow compared to when each truck travels alone; this is thanks to better use of road space, less lane changes, synchronous acceleration and deceleration among the trucks and less sudden braking. All the previous is possible thanks to the autonomous functionality and the V2V communication of the trucks. The ultimate goal is to form as many truck platoons per day as possible, but that presents a big number of difficulties and challenges. Competition among carriers, lack of V2I communication, absence of traffic prediction based trajectory planning, etc. all delay the accomplishment of this ultimate goal. This thesis provides control schemes and control-oriented models that aim to improve the usage of networks and infrastructure, reduce traffic congestion, and ultimately minimize travel times and energy/fuel consumption. To achieve these goals, the research presents a real-time planning framework for platoon coordination decisions based on traffic prediction, enabling autonomous vehicle groups to dynamically adapt to changing road conditions while optimizing efficiency. Additionally, a control-oriented highway traffic model is introduced, incorporating multiple clusters of connected autonomous vehicles to better capture the interactions between AVs and HVs and enhance traffic management strategies. Furthermore, the thesis addresses the minimization of energy consumption for electric autonomous vehicles by formulating optimal trajectory planning methods that leverage regenerative braking, ensuring sustainable and energy-efficient operations. By integrating these approaches, the proposed methodologies facilitate the safe and efficient coexistence of AVs and HVs while promoting sustainability in modern transportation systems.